With all the information out there, it can be difficult to even know how to understand and implement AI.
In my role on the AI Marketing Solutions (AIMS) team at HubSpot, I’ve seen first-hand how integrating AI can directly improve campaign metrics. My proudest moment so far was when I launched an AI-personalized email writing workflow that saw a 38% lift in overall CTR (click-through-rate) compared to our traditional campaigns.
Now, you might be wondering, how can you get similar results, or better?
7 Essential AI Insights for Marketers in 2024
1. Become a prompt master.
“With where AI is currently at, knowing how to craft winning prompts that get the results you need is really valuable,” says Matt Wolfe, an AI reporter and analyst whose YouTube channel has nearly half-a-million subscribers.
Let me give you a concrete illustration of what Wolfe means and how to apply it. Here’s how to upgrade your AI prompts:
- Be specific and goal-oriented: Define your intentions clearly.
- Consider your context and persona: How should the AI model approach the problem? As an experienced email marketer or a creative copywriter?
- Let’s talk output: Envision what the end result should look like.
“Act as a seasoned marketing blogger. Produce outputs a seasoned marketing blogger would produce. Craft an engaging, conversational post detailing five major marketing pitfalls specifically for IT companies, explaining their negative ramifications, aiming for 450 words.“
2. Develop a workflow.
Once you’ve gotten prompting down, then the real fun begins. Let’s break this down in a “day in the life” of an AI-enabled marketer.
- 8:30 AM: Review performance metrics from previous campaigns; ChatGPT generates a summary of the data, highlighting metrics and making reporting a breeze.
- 9:15 AM: Dive into creative mode. Direct AI to generate and refine email subject lines, and then shortlist the high-potential hitters for A/B testing.
- 11:00 AM: Use ChatGPT to organize a messy pile of customer feedback, categorizing comments for easy team review and response strategies.
- 1:00 PM: Ask ChatGPT about blog topic ideas, considering previous content we’ve made and don’t want to replicate.
- 2:30 PM: Work with AI to develop blog post titles and outlines, tapping into SEO insights, specifying keywords, and integrating current trends to maximize reach.
- 3:45 PM: Direct AI to mockup some social media posts, later refining the copy to inject brand voice and personality before scheduling.
- 4:30 PM: Conclude the day by having AI clean up and format a client presentation, ensuring clarity and professional polish.
You might be wondering, how do I tactically turn AI-learnings into marketing performance?
Here’s how I transformed our emails: I used a mix of AI-driven audience segmentation and RAG (retrieval augmented generation) techniques to write the most meaningful emails possible, capturing the 38% improvement in CTR that I mentioned earlier.
This bump is more than a number, though: It’s really about driving value. My A/B test didn’t just crown a winner; it reassured me that meaningfully knowing and addressing my audience leads to groundbreaking engagement. It’s proof that with personalization, precision pays off.
But here’s the quiet part: for AI-driven personalization to work, it needs to merge seamlessly with tried-and-true marketing strategies. AI isn’t something you can just plug in and supercharge your marketing. You have to be intentional, and even then, sometimes AI might not make a huge lift. Part of being an early adopter is being okay with failure.
3. Test all the tools.
HubSpot’s CTO, Dharmesh Shah, couldn’t put it better:
“Marketers need to recognize how far A.I. has come — and how fast it is moving,” says Shah.
“Before deciding that AI cannot help with a given task, try it. It’s possible it couldn’t be done a few months ago, but now it can.”
AI breakthroughs are accelerating. Think about a feature as simple as web-browsing. In the early days of ChatGPT, it was a well-known limit for AI. Now, you can ask ChatGPT almost any question and instruct the model to browse, and off it’ll go (as long as you have a Plus account).
He mentions the company Conveyor, which is working to narrow the latency time between when you ask a chatbot a question and when it responds. Latency is so important in chatbots because it dictates the entire user experience. It’s now possible to deliver nearly seamless, real-time, personal engagement.
Across the marketing world, there’s an AI tool with your brand’s name on it, waiting to tackle your specific challenges. Whether you’re managing customer communications, seeking insight from market data, or crafting visuals that capture and convert, AI could be the ace up your sleeve.
4. Use AI as your partner, not your replacement.
AI has already proven to be a strong contributor across marketing workflows.
Ramon Berrios, host of DTC Pod, has integrated AI into most of his company’s marketing. He uses his tool, castmagic.io, to produce their weekly industry newsletter, automate podcast production work, run social media, and record conversations with thought leaders.
Keep in mind, AI tools are here to back you up, not take over. With AI churning out endless streams of text, we might see a real flood of low-quality, spammy, AI-generated content.
As marketers, we know that high-quality content will always be the key to marketing success, and this is truer yet in the age of AI. Some might try to play the system, but the truth is that engaging and captivating content will always win in the long run.
Some of my tips to identify AI-generated content are as follows:
- Originality: Does the content offer unique perspectives not seen elsewhere?
- Insightfulness: Is it educated by in-depth knowledge, providing value that exceeds mere information regurgitation?
- Brand Alignment: does the content have a distinct brand voice? A distinct tone? Or does it sound like that ‘generic’ ChatGPT tone that’s becoming all too common?
As AI evolves, we’re going to have to figure out which marketing tasks are enhanced by AI and which ones definitely require a human touch. But along the way, we’re going to learn a ton, see some massive breakthroughs in what’s possible, and discover how to drive even more value for customers.
5. Look out for your brand.
Brand safety takes precedence in this new era. At HubSpot, we’ve ingrained checks to avoid any inappropriate or unintended interactions.
“Data importing can be a concern and a security breach for any organization. Make sure to really understand the terms and conditions of an AI platform before you upload any company-sensitive information and data,” says Berrios.
Additionally, make sure the use of AI is communicated transparently and effectively across your organization. Berrios emphasizes the importance of brand and tone alignment, especially when team members are using different AI platforms whose models don’t learn from each other.
6. Integrate intentionally.
While it might be tempting to go full-speed ahead in AI, there are inherent risks.
Here are my five AI no-go’s:
- Data Gathering Without Consent — Don’t be a privacy pirate.
- Misleading Big Data Expectations — AI isn’t your crystal ball.
- AI Tools Without Clean Data — Garbage in, garbage out.
- Ignoring AI Source Verification — Fact-check or face the music.
- Neglecting Ethics in AI Use — Skirt morals, invite scandals.
There’s a prevalent yet flawed notion that AI can handle vast and unfocused datasets to drive insightful conclusions. However, the reality is that AI excels when led with specificity and purpose. It demands data hygiene and ethical data use — these are cornerstones of HubSpot’s AI strategy.
- Transparency in Data Collection: Clearly itemize the types of personal information your company collects, spanning from the obvious — names, email addresses — to the less considered, like browsing behavior or device information. Leave no stone or data point unturned.
- User Consent and Control: Spell out how users can give, manage, and revoke consent for data use. Include how to access, correct, or delete their personal information.
- Data Security and Sharing: Articulate how you keep user data under lock and digital key. Be clear on who’s in your data-sharing circle and why.
During your creative process, beware of misleading information, as well as copyright infringement. Use tools that train their models on clean data. Double-check and diversify your sources to avoid misinformation.
This video on AI Ethics is a great place to find common AI pitfalls and solutions.
7. Stay human.
AI may not be human (yet), but luckily, you are. As we move into 2024, the goal is to further discover the potential and limitations of AI in marketing, making sure that our applications remain powerful, meaningful, and value-driven.
“When a machine can build any product in the world, the product that’s going to win is going to have a real point of view, and the people marketing it…are going to have a real point of view. [Central to everything], is believing deeply in something,” says Kipp Bodnar, CMO at HubSpot.
AI is a work in progress, which means it is still disproportionately biased against minorities, and may offer perspectives that aren’t informed by empathy and critical thinking. Ironically, in the age of machines, your humanity is what will truly make the difference.
The key to winning at AI is to be strategic, intentional, and vigilant. AI is a tool to be wielded wisely — to amplify our potential, not to overshadow it. It’s about enriching brand experiences, while still reflecting the warmth, sincerity, and value-focused care of our human marketers.